Based on a handful of results from the Beta project, I estimate the Enhanced app (v5.10) to average about fifteen times the CPU time taken by a current optimized app (Team MacNN's) on my G4/733. ... Compared to the current standard app that would probably be closer to five times as long.

On this pc enhanced, ver 5.11, only takes 2.5 times longer than standard app, at most common a_r = 0.422.

Well, 2.5 is certainly closer to 5 than it is to 15. ;p

---

And if the "enhanced" version of BOINC software "will slow everyone down"

Not BOINC, but Seti. Seti-Enhanced will slow down the crunching time. Albeit no longer by the amounts previously told (10 - 30 times), you won't be able to crunch them in (under) two hours either.

Just in case anyone gets the wrong idea, the "slowing down" isn't a matter of figuratively applying brakes; the SETI Enhanced app does a much more thorough analysis than the current ones; it should be able to find signals that would be missed by earlier versions.

Based on a handful of results from the Beta project, I estimate the Enhanced app (v5.10) to average about fifteen times the CPU time taken by a current optimized app (Team MacNN's) on my G4/733. (The developers have been doing some optimization themselves; the factor for v5.05 was more like fifty! The version that gets released may be even more streamlined.) Compared to the current standard app that would probably be closer to five times as long.

On this pc enhanced, ver 5.11, only takes 2.5 times longer than standard app, at most common a_r = 0.422.

Andy

And if the "enhanced" version of BOINC software "will slow everyone down"

Not BOINC, but Seti. Seti-Enhanced will slow down the crunching time. Albeit no longer by the amounts previously told (10 - 30 times), you won't be able to crunch them in (under) two hours either.

Just in case anyone gets the wrong idea, the "slowing down" isn't a matter of figuratively applying brakes; the SETI Enhanced app does a much more thorough analysis than the current ones; it should be able to find signals that would be missed by earlier versions.

Based on a handful of results from the Beta project, I estimate the Enhanced app (v5.10) to average about fifteen times the CPU time taken by a current optimized app (Team MacNN's) on my G4/733. (The developers have been doing some optimization themselves; the factor for v5.05 was more like fifty! The version that gets released may be even more streamlined.) Compared to the current standard app that would probably be closer to five times as long.

---

Enhanced is not slower. It does a more in depth analysis of the work unit, so it takes longer to produce the results. This is a good thing. It increases the chances of finding the signal that we are looking for.
:)

---

98SE XP2500+ @ 2.1 GHz Boinc v5.8.8

And God said"Let there be light."But then the program crashed because he was trying to access the 'light' property of a NULL universe pointer.

And if the "enhanced" version of BOINC software "will slow everyone down"

Not BOINC, but Seti. Seti-Enhanced will slow down the crunching time. Albeit no longer by the amounts previously told (10 - 30 times), you won't be able to crunch them in (under) two hours either.

---

If and when the "enhanced" version is released it will slow everyone down and hopefully these weekly down sessions will not need to be done.

The need for doing backups will not go-away by the introduction of Seti_Enhanced...

To not go down for weekly backup-outages, they needs a new server powerful enough to act as "slave"-database, so instead of taking-down the "live" database they can instead take-down the slave-database and backup from this.

There will of course be other non-backup-related outages still happening even after someone has either donated a new server, or enough money is donated so they themselves can buy a new server...

I know how you feel.
I remember a very cool submarine game I used to play. It was on ONE floppy (single sided) and was played on my Atari 520ST which also used ONE floppy disc for the ENTIRE OS.

Yea, back then I was writing my own version of the StarTreak game and running it on a home built 8080 machine with only 4K of memory.

Do you really know how I feel. Just let me tell you.

I recently had need to use some software program on old engine controllers (that connects by an ASYNC adpater and sold by a company that has been out of business for 15 years) in a remote flooded location to solve several problems in a location you would not want to take an expensive computer or one you cared about working afterward. No one was interested in spending any more time and effort trying to find the problems, and I was only interested for personal reasons. Only several programs costing $350 or more would do the job. They all required Windows 9x/XP and at least a P-II processor with 128MB to run, had 40MB or larger install programs. The lower cost programs and smaller programs available were not able to connect with and do what I needed, as non standard ASYNC drivers are required. Having limited personal funds for this one time use, and only a very old laptop running DOS on a 4 MB disk drive to take, I had to settle for a very hard to use and unreliable but free program that was only 140K in size and would run under DOS, and only recorded the raw data. Took several trips to collect data, come back and sort out what the bits mean, go back for more, then back to use the program to "fix" the problems. But it did the job at little cost to me, just extra time, trips and effort. I don't expect to get compensated for my effort.

Got so pissed off about it all, that I am writing my own program in case I ever have reason to do this dumb thing again. It now does everything the free program did and more, except very reliable and faster, automatically without a lot of manual hex input. Will run on an Intel 8088 class processor (PC/XT) with only 64K of memory with just a boot diskette if necessary, (a machine that I can take and leave there) and has its own ASYNC chip drivers, will display the raw data with its own display drivers, and detect common communications error problems that even the expensive programs will not do. Oh, its up to a whopping 2523 bytes in size already. I have to add about 1K more code to sort out the raw data, parse it into control blocks and display them in a readable form, so I can see what is being collected in real time. That will make it much more usable. Once that is done, I might add the other 40K or so of code and display data to actually analise and display the data and what it means in real time. I would then be able to do most of what the expensive programs could do, do it quicker and using a much cheaper and more disposable computer. And can let someone else younger wade several miles through three foot deep swamp and muck, and use a cell phone to ask what to do next. The best part, with a few more modifications, the program will be usable for other things.

Faster computers are necessary to make up for the lower quality programs that keep getting slower and more overbloated.

I know how you feel.
I remember a very cool submarine game I used to play. It was on ONE floppy (single sided) and was played on my Atari 520ST which also used ONE floppy disc for the ENTIRE OS.

That SUCKS A LOT! Everu Tuesday will be this... =/

If and when the "enhanced" version is released it will slow everyone down and hopefully these weekly down sessions will not need to be done.

I am to old and impatient to be slowed down!

I have dealt with computers for 42 years.
They keep getting faster.
I keep getting slower and more impatient.
Faster computers are necessary to make up for the lower quality programs that keep getting slower and more overbloated.

Computers are fast, networks are reasonably fast, so why should we, mere slow carbon based computers, have to wait on them. After all they are our servants, not the other way around. And if the "enhanced" version of BOINC software "will slow everyone down", then it will be time for everyone to evaluate their commitment of their limited and valuable time to Setiathome and BOINC slow programs. And a project that will not address their slow performance problems in a professional manor, and instead blame them on their customers behavior.

However, before I, and others, blame todays slow problems all on BOINC and the setiathome server software, I at least need to check our facts. And in so doing, I have found that today, unlike most of the times when recoveries are very slow, there is something else wrong beyond our control.

Seems there is a bad case of network congestion and delay being experienced by all the network routers in the LAX/LAS exchanges. Twice normal delay and ping times through those routers with a high (up to 15% or so) dropped packet problems. Saw those same delays on my ISP's routers in the LA area, as well as others including the CogentCo routers that all uploads and downloads go through.

So this time we have to blame it on some sort of "network congestion" problem.

That SUCKS A LOT! Everu Tuesday will be this... =/

If and when the "enhanced" version is released it will slow everyone down and hopefully these weekly down sessions will not need to be done.

---

Boinc....Boinc....Boinc....Boinc....

Well, it just went through. I think you were good luck. or maybe it's the squeeky wheel thing.

At any rate, happy munching, crunching and all that....

Toodles!!

---

I just checked the server page and it looks like everything is OK. Why am I getting the message 'couldn't connect to hostname setiboincdata.ssl.berkely.edu??

Merci, as usual ;-)

4/18/06 9:38:15 PM|SETI@home|Started upload of 01jn01aa.23716.3393.11064.1.164_3_0
4/18/06 9:39:23 PM||Couldn't connect to hostname [setiboincdata.ssl.berkeley.edu]
4/18/06 9:39:23 PM|SETI@home|Temporarily failed upload of 01jn01aa.23716.3393.11064.1.164_3_0: system I/O
4/18/06 9:39:23 PM|SETI@home|Backing off 2 hours, 57 minutes, and 45 seconds on upload of file 01jn01aa.23716.3393.11064.1.164_3_0

It happens from time to time. Either the server is backed up with too much traffic or there is simply a communication break down. BIONIC will automatically back off for a set amount of time and try reporting at a later time.

That SUCKS A LOT! Everu Tuesday will be this... =/

I don't see any difference from it being on a wednesday as opposed to it being a Tuesday now. Its nothing new. Its making it easier on the UCB staff to do everything on Tuesday, since EVERYONE will be there instead of half on wednesday. I'm all for it.

And if you don't like the fact that they have to bring the servers down every week, join another project, set the resources low, give it a connect every day setting of .1, and your good to go. :-) Or, if you just crunch strictly SETI, then set your connect days to 10, or less depending on how fast of a machine you have.

Just some thoughts on how to get around this. :-) Happy crunching.

---

I just checked the server page and it looks like everything is OK. Why am I getting the message 'couldn't connect to hostname setiboincdata.ssl.berkely.edu??

Merci, as usual ;-)

4/18/06 9:38:15 PM|SETI@home|Started upload of 01jn01aa.23716.3393.11064.1.164_3_0
4/18/06 9:39:23 PM||Couldn't connect to hostname [setiboincdata.ssl.berkeley.edu]
4/18/06 9:39:23 PM|SETI@home|Temporarily failed upload of 01jn01aa.23716.3393.11064.1.164_3_0: system I/O
4/18/06 9:39:23 PM|SETI@home|Backing off 2 hours, 57 minutes, and 45 seconds on upload of file 01jn01aa.23716.3393.11064.1.164_3_0

It happens from time to time. Either the server is backed up with too much traffic or there is simply a communication break down. BIONIC will automatically back off for a set amount of time and try reporting at a later time.

That SUCKS A LOT! Everu Tuesday will be this... =/

---

My 4 hosts are also backed up with uploading timeouts. The server page shows all is okay but the homepage says that the project weekly backup was moved to Tuesday from Wednesday. The upload timeouts are probably just normal server connection issues after recovering from the ~4 hour outage.

---

I just checked the server page and it looks like everything is OK. Why am I getting the message 'couldn't connect to hostname setiboincdata.ssl.berkely.edu??

Merci, as usual ;-)

4/18/06 9:38:15 PM|SETI@home|Started upload of 01jn01aa.23716.3393.11064.1.164_3_0
4/18/06 9:39:23 PM||Couldn't connect to hostname [setiboincdata.ssl.berkeley.edu]
4/18/06 9:39:23 PM|SETI@home|Temporarily failed upload of 01jn01aa.23716.3393.11064.1.164_3_0: system I/O
4/18/06 9:39:23 PM|SETI@home|Backing off 2 hours, 57 minutes, and 45 seconds on upload of file 01jn01aa.23716.3393.11064.1.164_3_0

It happens from time to time. Either the server is backed up with too much traffic or there is simply a communication break down. BIONIC will automatically back off for a set amount of time and try reporting at a later time.

---

I just checked the server page and it looks like everything is OK. Why am I getting the message 'couldn't connect to hostname setiboincdata.ssl.berkely.edu??

Merci, as usual ;-)

4/18/06 9:38:15 PM|SETI@home|Started upload of 01jn01aa.23716.3393.11064.1.164_3_0
4/18/06 9:39:23 PM||Couldn't connect to hostname [setiboincdata.ssl.berkeley.edu]
4/18/06 9:39:23 PM|SETI@home|Temporarily failed upload of 01jn01aa.23716.3393.11064.1.164_3_0: system I/O
4/18/06 9:39:23 PM|SETI@home|Backing off 2 hours, 57 minutes, and 45 seconds on upload of file 01jn01aa.23716.3393.11064.1.164_3_0

---